CN116522183B - Sub-class center point feature extraction method, system and equipment based on seed clustering - Google Patents
Sub-class center point feature extraction method, system and equipment based on seed clustering Download PDFInfo
- Publication number
- CN116522183B CN116522183B CN202310048970.8A CN202310048970A CN116522183B CN 116522183 B CN116522183 B CN 116522183B CN 202310048970 A CN202310048970 A CN 202310048970A CN 116522183 B CN116522183 B CN 116522183B
- Authority
- CN
- China
- Prior art keywords
- singular
- subclass
- class
- data set
- subclasses
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000605 extraction Methods 0.000 title claims abstract description 46
- 238000005520 cutting process Methods 0.000 claims abstract description 63
- 230000007717 exclusion Effects 0.000 claims abstract description 51
- 238000000034 method Methods 0.000 claims description 48
- 238000004364 calculation method Methods 0.000 claims description 23
- 238000003860 storage Methods 0.000 claims description 13
- 238000012163 sequencing technique Methods 0.000 claims description 8
- 238000012217 deletion Methods 0.000 claims description 4
- 230000037430 deletion Effects 0.000 claims description 4
- 230000009191 jumping Effects 0.000 claims description 3
- 230000011218 segmentation Effects 0.000 claims description 3
- 101100391182 Dictyostelium discoideum forI gene Proteins 0.000 claims description 2
- 238000005070 sampling Methods 0.000 description 10
- 230000008569 process Effects 0.000 description 7
- 238000004220 aggregation Methods 0.000 description 6
- 238000007726 management method Methods 0.000 description 6
- 238000012545 processing Methods 0.000 description 5
- 239000012297 crystallization seed Substances 0.000 description 4
- 238000004422 calculation algorithm Methods 0.000 description 3
- 238000004891 communication Methods 0.000 description 3
- 239000000284 extract Substances 0.000 description 3
- 238000009395 breeding Methods 0.000 description 2
- 230000001488 breeding effect Effects 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 238000009826 distribution Methods 0.000 description 2
- 230000008030 elimination Effects 0.000 description 2
- 238000003379 elimination reaction Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000007619 statistical method Methods 0.000 description 2
- 238000006467 substitution reaction Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000010295 mobile communication Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 230000007723 transport mechanism Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2337—Non-hierarchical techniques using fuzzy logic, i.e. fuzzy clustering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/10—Pre-processing; Data cleansing
- G06F18/15—Statistical pre-processing, e.g. techniques for normalisation or restoring missing data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/2433—Single-class perspective, e.g. one-against-all classification; Novelty detection; Outlier detection
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Fuzzy Systems (AREA)
- Software Systems (AREA)
- Mathematical Physics (AREA)
- Automation & Control Theory (AREA)
- Probability & Statistics with Applications (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a seed clustering-based sub-class center point feature extraction method, a seed clustering-based sub-class center point feature extraction system and seed clustering-based sub-class center point feature extraction equipment, wherein the number of seed data sets and extracted point feature samples is obtained; clustering the seed data set; dividing the clustered data set into singular large subclasses, singular small subclasses and singular intermediate classes; calculating a cutting threshold value of the singular large subclass, an exclusion threshold value of the singular small subclass and an exclusion threshold value of the singular middle class; cutting the singular large subclass according to a cutting threshold value of the singular large subclass to obtain a first data set; according to the exclusion threshold value of the singular small subclass, the singular small subclass is excluded, and a second data set is obtained; removing the singular intermediate class according to the removal threshold value of the singular intermediate class to obtain a third data set; and merging the first data set, the second data set and the third data set to obtain a merged data set, and extracting the center point characteristic corresponding to each subclass in the merged data set. The invention can improve the scientificity and the accuracy of the comprehensive judgment of seeds.
Description
Technical Field
The invention relates to the technical field of feature extraction, in particular to a method, a system and equipment for extracting sub-class center point features based on seed clustering.
Background
In the processes of crop breeding, seed production and seed processing, the characteristic parameters of the same batch of seeds such as grain type, grain weight and the like are required to be obtained, representative seed samples are generally manually sampled according to a certain proportion, the characteristic parameters of the seed samples such as grain type, grain weight and the like are obtained, and the characteristic parameters of the artificial sampling samples are adopted as the characteristic parameters of the batch of seeds, so that the same batch of samples are often inspected and statistically analyzed in a manual sampling mode, the quality, purity, discrete condition and the like of the batch of seeds are judged, certain continuity of the seed characteristic parameters is considered, normal distribution is met, and the seed characteristic parameters are influenced by manual sampling, variety, year and cultivation management to have certain fluctuation. The traditional statistical analysis based on artificial sampling is difficult to embody the characteristic data parameters of the whole seed lot due to error factors such as artificial sampling, variety, year and cultivation management, and the dynamic change of the characteristic parameters of the seed lot due to variety, year and cultivation management mode cannot be represented. Therefore, scientific and accurate characteristic data cannot be provided for seed quality judgment.
Disclosure of Invention
The present invention aims to solve at least one of the technical problems existing in the prior art. Therefore, the invention provides the subclass center point feature extraction method, the subclass center point feature extraction system and the subclass center point feature extraction equipment based on seed clustering, which can improve the scientificity and the accuracy of seed comprehensive judgment.
In a first aspect, an embodiment of the present invention provides a method for extracting a feature of a sub-class center point based on seed clustering, where the method for extracting a feature of a sub-class center point based on seed clustering includes:
acquiring a seed data set and the number of extracted point characteristic samples;
clustering the seed data set according to the number of the extracted point characteristic samples to obtain a clustered data set;
dividing the clustered data set into singular large subclasses, singular small subclasses and singular intermediate classes;
according to the set of the singular large subclass, the singular small subclass and the singular intermediate class, calculating a cutting threshold of the singular large subclass, an exclusion threshold of the singular small subclass and an exclusion threshold of the singular intermediate class;
cutting the singular large subclass according to a cutting threshold value of the singular large subclass to obtain a first data set;
removing the singular small subclasses according to the removal threshold value of the singular small subclasses to obtain a second data set;
Removing the singular intermediate class according to the removal threshold value of the singular intermediate class to obtain a third data set;
and merging the first data set, the second data set and the third data set to obtain a merged data set, and extracting the center point characteristic corresponding to each subclass in the merged data set.
Compared with the prior art, the first aspect of the invention has the following beneficial effects:
according to the method, the seed data set and the number of the extracted point characteristic samples are obtained, and the seed data set is clustered according to the number of the extracted point characteristic samples, so that the singular subclasses are easier to find, and the subsequent singular subclasses are convenient to process; calculating a cutting threshold value of the singular large subclass, an exclusion threshold value of the singular small subclass and an exclusion threshold value of the singular intermediate class, cutting the singular large subclass according to the cutting threshold value of the singular large subclass, excluding the singular small subclass according to the exclusion threshold value of the singular small subclass, excluding the singular intermediate class according to the exclusion threshold value of the singular intermediate class, and cutting and excluding the singular subclass so as to exclude the influence of the singular subclass on the whole data set, thereby ensuring the scientificity and the accuracy of each subclass; the first data set, the second data set and the third data set are combined to obtain a combined data set, the central point characteristic corresponding to each sub-class in the combined data set is extracted, the scientific and accurate characteristic data are provided for seed quality judgment through extracting the central point characteristic parameters corresponding to each sub-class, the calculation process of seed comprehensive judgment is simplified, the accuracy of seed comprehensive judgment is improved, the calculation complexity is lower, and the realizability is high.
According to some embodiments of the present invention, before the classifying the clustered dataset into the singular large subclass, the singular small subclass and the singular middle subclass, the seed-clustering-based subclass center point feature extraction method further includes:
presetting a threshold value of the minimum sample number in each subclass；
Computing a set of sample numbers within each sub-class of the clustered datasetThe method comprises the following steps:
wherein ,the representation is->Operators of the number of elements in the subclass;
calculation of25% quantile->：
wherein ,representing a quantile extraction operator;
calculation of75% quantile->：
；
Calculation ofMean point of->：
；
Calculating the judging threshold value of the singular large subclassThe method comprises the following steps: />；
Calculating the judging threshold value of the singular small subclassThe method comprises the following steps: />。
According to some embodiments of the invention, the classifying the clustered data sets into the singular large subclass, the singular small subclass, and the singular intermediate class includes:
in the clustered dataset, ifThen the i-th subclass->Is judged as singular large subclasses and each set of singular large subclasses is expressed as +.>, wherein ,representation set->The number of subclasses of (3);
in the clustered dataset, ifThen the i-th subclass->Is judged as singular small subclasses and each set of singular small subclasses is expressed as +. >, wherein ,representation set->The number of subclasses of (3);
in the clustered dataset, ifAnd->Then the i-th subclass->Is judged as singular intermediate classes and each set of singular intermediate classes is expressed as, wherein ,/>Representation set->Is a subclass of (c).
According to some embodiments of the invention, the calculating the cutting threshold of the singular large subclass, the eliminating threshold of the singular small subclass and the eliminating threshold of the singular intermediate class according to the set of the singular large subclass, the singular small subclass and the singular intermediate class comprises:
for the setThe ith subclass of (2)>Calculating the subclass +.>Divided into +.>：
wherein ,representing a down-rounding operator;
the number of cuts per subclassMake up the collection->；
Counting the number of deletions as, wherein ,/>Representing the number of samples of the extracted point feature,representing the clustered numbersThe number of subclasses in the dataset;
calculating the intermediate quantity value as；
Calculating a cutting threshold of the singular large subclassThe method comprises the following steps:
wherein ,represents a summation operator;
calculating an exclusion threshold for the singular subclassThe method comprises the following steps:
；
calculating an exclusion threshold for the singular intermediate classThe method comprises the following steps:
。
according to some embodiments of the invention, the step of cutting the singular large subclass according to a cutting threshold of the singular large subclass to obtain a first data set includes:
For the setCutting each subclass in (3), if the cutting threshold of the singular large subclass isThen it is made by the followingThe steps are divided:
step S111, for the setSetting i as the set +.>Sequence number of each subclass in (a), and；
step S112, ifTaking the set->Is>Individual subset->And obtain the subsetData dimension +.>；
Step S113, calculating the subsetIs>Distance of 90% quantiles from 10% quantiles of the individual dimension data；
Step S114, obtaining the maximum distance value as，/>Representation->Corresponding dimension serial numbers;
step S115, at the firstIn dimension, sorting is carried out according to the data size, and the sorted set is that；
Step S116, forIs added to the one-dimensional data before ordering>The data of the same column are combined in the column direction to form a new +.>Dimensional data set；
Step S117, according to the setThe corresponding division number->For->Cutting sequentially to obtain +.>, wherein ,/>Representation->And let +.>Jump to step S112;
step S118, ifThe cut data is collected as；
If the cutting threshold of the singular large subclassThen->, wherein ,/>Representing the absolute value taking operator.
According to some embodiments of the invention, the excluding the singular small subclass according to the excluding threshold of the singular small subclass, obtaining the second data set includes:
step S121, if the exclusion threshold of the singular small subclassLet->Jump to step S125;
step S122, eliminating threshold value of the singular small subclassAnd is also provided withThen->Jump toStep S124;
step S123, ifOrder in principleWill->According to the set->Sequencing the number of internal samples of the sub-set to obtain a sequenced set +.>And (3) removing +.>Middle frontA subset of individuals, get the set->，/>Representation->Excluding the number of subsets;
step S124, collecting the setThe empty subset of the Chinese is excluded to obtain the set +.>；
Step S125, save set。
According to some embodiments of the invention, the excluding the singular intermediate class according to the excluding threshold of the singular intermediate class, obtaining a third data set includes:
step S131, if the singular intermediate class is eliminatedLet->Jump to step S134;
step S132, if the singular intermediate class is eliminatedAnd is also provided with
Then->Jump to step S134;
step S133, if Order in principleWill->According to the set->Sequencing the number of internal samples of the sub-set to obtain a sequenced set +.>And (3) removing +.>Middle frontA subset of individuals, get the set->Let->；
Step S134, save collection。
In a second aspect, an embodiment of the present invention further provides a system for extracting a feature of a sub-class center point based on seed clustering, where the system for extracting a feature of a sub-class center point based on seed clustering includes:
the data acquisition unit is used for acquiring a seed data set and the number of the extracted point characteristic samples;
the data clustering unit is used for clustering the seed data set according to the number of the extracted point characteristic samples to obtain a clustered data set;
the subclass dividing unit is used for dividing the clustered data set into a singular large subclass, a singular small subclass and a singular middle class;
the threshold calculating unit is used for calculating a cutting threshold of the singular large subclass, an exclusion threshold of the singular small subclass and an exclusion threshold of the singular intermediate class according to the set of the singular large subclass, the singular small subclass and the singular intermediate class;
the subclass cutting unit is used for cutting the singular large subclass according to a cutting threshold value of the singular large subclass to obtain a first data set;
The first excluding unit is used for excluding the singular small subclasses according to the excluding threshold value of the singular small subclasses to obtain a second data set;
the second excluding unit is used for excluding the singular intermediate class according to the excluding threshold value of the singular intermediate class to obtain a third data set;
and the feature extraction unit is used for merging the first data set, the second data set and the third data set to obtain a merged data set, and extracting the center point feature corresponding to each subclass in the merged data set.
In a third aspect, an embodiment of the present invention further provides a seed cluster-based sub-class center point feature extraction device, including at least one control processor and a memory communicatively connected to the at least one control processor; the memory stores instructions executable by the at least one control processor to enable the at least one control processor to perform a seed cluster based sub-class center feature extraction method as described above.
In a fourth aspect, embodiments of the present invention further provide a computer-readable storage medium storing computer-executable instructions for causing a computer to perform a seed-cluster-based sub-class center point feature extraction method as described above.
It is to be understood that the advantages of the second to fourth aspects compared with the related art are the same as those of the first aspect compared with the related art, and reference may be made to the related description in the first aspect, which is not repeated herein.
Drawings
The foregoing and/or additional aspects and advantages of the invention will become apparent and may be better understood from the following description of embodiments taken in conjunction with the accompanying drawings in which:
FIG. 1 is a flowchart of a method for extracting features of sub-class center points based on seed clustering according to an embodiment of the present invention;
fig. 2 is a block diagram of a seed-cluster-based sub-class center feature extraction system according to an embodiment of the present invention.
Detailed Description
Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to like or similar elements or elements having like or similar functions throughout. The embodiments described below by referring to the drawings are illustrative only and are not to be construed as limiting the invention.
In the description of the present invention, the description of first, second, etc. is for the purpose of distinguishing between technical features only and should not be construed as indicating or implying a relative importance or implicitly indicating the number of technical features indicated or implicitly indicating the precedence of the technical features indicated.
In the description of the present invention, it should be understood that the direction or positional relationship indicated with respect to the description of the orientation, such as up, down, etc., is based on the direction or positional relationship shown in the drawings, is merely for convenience of describing the present invention and simplifying the description, and does not indicate or imply that the apparatus or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and thus should not be construed as limiting the present invention.
In the description of the present invention, unless explicitly defined otherwise, terms such as arrangement, installation, connection, etc. should be construed broadly and the specific meaning of the terms in the present invention can be determined reasonably by a person skilled in the art in combination with the specific content of the technical solution.
In the processes of crop breeding, seed production and seed processing, the characteristic parameters of the same batch of seeds such as grain type, grain weight and the like are required to be obtained, representative seed samples are generally manually sampled according to a certain proportion, the characteristic parameters of the seed samples such as grain type, grain weight and the like are obtained, and the characteristic parameters of the artificial sampling samples are adopted as the characteristic parameters of the batch of seeds, so that the same batch of samples are often inspected and statistically analyzed in a manual sampling mode, the quality, purity, discrete condition and the like of the batch of seeds are judged, certain continuity of the seed characteristic parameters is considered, normal distribution is met, and the seed characteristic parameters are influenced by manual sampling, variety, year and cultivation management to have certain fluctuation. The traditional statistical analysis based on artificial sampling is difficult to embody the characteristic data parameters of the whole seed lot due to error factors such as artificial sampling, variety, year and cultivation management, and the dynamic change of the characteristic parameters of the seed lot due to variety, year and cultivation management mode cannot be represented. Therefore, scientific and accurate characteristic data cannot be provided for seed quality judgment.
According to the invention, the seed data set and the number of the extracted point characteristic samples are obtained, and the seed data set is clustered according to the number of the extracted point characteristic samples, so that the singular subclass is easier to find, and the subsequent singular subclass is convenient to process; calculating a cutting threshold value of the singular large subclass, an exclusion threshold value of the singular small subclass and an exclusion threshold value of the singular intermediate class, cutting the singular large subclass according to the cutting threshold value of the singular large subclass, excluding the singular small subclass according to the exclusion threshold value of the singular small subclass, excluding the singular intermediate class according to the exclusion threshold value of the singular intermediate class, and cutting and excluding the singular subclass so as to exclude the influence of the singular subclass on the whole data set, thereby ensuring the scientificity and the accuracy of each subclass; the first data set, the second data set and the third data set are combined to obtain a combined data set, the central point characteristic corresponding to each sub-class in the combined data set is extracted, the scientific and accurate characteristic data are provided for seed quality judgment through extracting the central point characteristic parameters corresponding to each sub-class, the calculation process of seed comprehensive judgment is simplified, the accuracy of seed comprehensive judgment is improved, the calculation complexity is lower, and the realizability is high.
Referring to fig. 1, an embodiment of the present invention provides a method for extracting a feature of a sub-class center point based on seed clustering, where the method for extracting a feature of a sub-class center point based on seed clustering includes, but is not limited to, steps S100 to S800:
step S100, acquiring a seed data set and the number of extracted point characteristic samples;
step 200, clustering the seed data set according to the number of the extracted point characteristic samples to obtain a clustered data set;
step S300, dividing the clustered data set into a singular large subclass, a singular small subclass and a singular middle class;
step S400, calculating a cutting threshold value of the singular large subclass, an exclusion threshold value of the singular small subclass and an exclusion threshold value of the singular intermediate class according to the set of the singular large subclass, the singular small subclass and the singular intermediate class;
s500, cutting the singular large subclasses according to a cutting threshold value of the singular large subclasses to obtain a first data set;
step S600, eliminating the singular small subclasses according to an elimination threshold value of the singular small subclasses to obtain a second data set;
step S700, eliminating the singular intermediate class according to an elimination threshold value of the singular intermediate class to obtain a third data set;
step S800, merging the first data set, the second data set and the third data set to obtain a merged data set, and extracting the central point characteristics corresponding to each subclass in the merged data set.
In steps S100 to S800 of some embodiments, in order to find singular subclasses more easily, facilitate processing of subsequent singular subclasses, first, the seed dataset is clustered according to the number of extracted point feature samples by obtaining the seed dataset and the number of extracted point feature samples; in order to ensure the scientificity and the accuracy of each subclass, a cutting threshold value of the singular large subclass, an exclusion threshold value of the singular small subclass and an exclusion threshold value of the singular intermediate class are calculated, the singular large subclass is cut according to the cutting threshold value of the singular large subclass, the singular small subclass is excluded according to the exclusion threshold value of the singular small subclass, and the singular intermediate class is excluded according to the exclusion threshold value of the singular intermediate class; in order to provide scientific and accurate characteristic data for seed quality judgment, simplify the calculation process of seed comprehensive judgment, improve the accuracy of seed comprehensive judgment, enable the calculation complexity to be lower and the realizability to be high, the application obtains a combined data set by combining the first data set, the second data set and the third data set, and extracts the center point characteristic corresponding to each subclass in the combined data set.
In some embodiments, before classifying the clustered dataset into the singular large subclass, the singular small subclass and the singular middle subclass, the seed-clustering-based subclass center point feature extraction method further comprises:
Presetting a threshold value of the minimum sample number in each subclass；
Computing a set of sample numbers within each sub-class in the clustered datasetThe method comprises the following steps:
wherein ,the representation is->Operators of the number of elements in the subclass;
calculation of25% quantile->：
wherein ,representing a quantile extraction operator;
calculation of75% quantile->：
；
Calculation ofMean point of->：
；
Calculating the judging threshold value of singular large subclassThe method comprises the following steps: />；
Calculating the judgment threshold value of singular subclassThe method comprises the following steps: />。
In this embodiment, the accuracy of dividing the singular large subclass, the singular small subclass and the singular intermediate class can be improved by the judgment threshold calculated in this embodiment.
In some embodiments, classifying the clustered data sets into a large singular subclass, a small singular subclass, and a middle singular class includes:
in the clustered dataset, ifThen the i-th subclass->Is judged as singular large subclasses and each set of singular large subclasses is expressed as +.>, wherein ,/>Representation set->The number of subclasses of (3);
in the clustered dataset, ifThen the i-th subclass->Is judged as singular small subclasses and each set of singular small subclasses is expressed as +.>, wherein ,/>Representation set->The number of subclasses of (3);
in the clustered dataset, if And->Then the i-th subclass->Is judged as singular intermediate classes and each set of singular intermediate classes is expressed as, wherein ,/>Representation set->Is a subclass of (c).
In the embodiment, the clustered data set is divided into the singular large subclass, the singular small subclass and the singular middle subclass, so that the singular subclass is conveniently removed, and the accuracy of the feature extraction of the central point can be improved.
In some embodiments, calculating the cut threshold for the big singular subclass, the exclude threshold for the small singular subclass, and the exclude threshold for the middle singular class from the set of the big singular subclass, the small singular subclass, and the middle singular class comprises:
for collectionsThe ith subclass of (2)>Calculator class->Divided into +.>：
wherein ,representing a down-rounding operator;
the number of cuts per subclassMake up the collection->；
Counting the number of deletions as, wherein ,/>Representing the number of samples of the extracted point feature,representing the number of subclasses in the clustered dataset;
calculating the intermediate quantity value as；
Calculating a cutting threshold of the singular large subclassThe method comprises the following steps:
wherein ,represents a summation operator;
calculating an exclusion threshold for singular subclassesThe method comprises the following steps:
；
calculating an exclusion threshold for singular intermediate classesThe method comprises the following steps:
。
in this embodiment, the accuracy of cutting the singular large subclass, excluding the singular small subclass, and excluding the singular intermediate class can be improved by the cutting threshold and the exclusion threshold calculated in this embodiment.
In some embodiments, the singular large subclass is cut according to a cut threshold of the singular large subclass to obtain a first data set, comprising:
pair aggregationCutting each subclass in (3), if the cutting threshold of the singular large subclass is +.>The segmentation is performed by the following steps:
step S111, pair aggregationSet i as set +.>Sequence number of each subclass in (a) and +.>；
Step S112, ifGet the collection->Is>Individual subset->And obtain subset->Data dimension +.>；
Step S113, calculating a subsetIs>Distance of 90% quantiles from 10% quantiles of the individual dimension data；
Step S114, obtaining the maximum distance value as，/>Representation->Corresponding dimension serial numbers;
step S115, at the firstIn dimension, sorting is carried out according to the data size, and the sorted set is that；
Step S116, forOne-dimensional data is added to the one-dimensional data before ordering>The data of the same column are combined in the column direction to form a new +.>Dimensional data set；
Step S117, according to the collectionThe corresponding division number->For->Cutting sequentially to obtain +.>, wherein ,/>Representation->And let +.>Jump to step S112;
step S118, ifThe cut data is collected as ；
Cutting threshold of singular large subclassThen->, wherein ,/>Representing the absolute value taking operator.
In this embodiment, the singular large subclass is cut according to the cutting threshold value of the singular large subclass, so that the accuracy of extracting the central point features can be improved.
In some embodiments, excluding the small singular subclasses according to an exclusion threshold of the small singular subclasses, obtaining the second data set comprises:
step S121, if the exclusion threshold of the singular subclassLet->Jump to step S125;
step S122, eliminating threshold value of singular small subclassAnd is also provided withThen->Jumping to step S124;
step S123, ifOrder in principleWill->According to the->Sequencing the number of internal samples of the sub-set to obtain a sequenced set +.>And (3) removing +.>Front middle>A subset of individuals, get the set->，/>Representation->Excluding the number of subsets;
step S124, collectingThe empty subset of the Chinese is excluded to obtain the set +.>；
Step S125, save set。
In this embodiment, the singular small subclass is excluded according to the exclusion threshold of the singular small subclass, so that the accuracy of extracting the central point features can be improved.
In some embodiments, excluding the singular intermediate class according to an exclusion threshold of the singular intermediate class, obtaining the third dataset comprises:
Step S131, if the singular intermediate class is excluded from the thresholdLet->Jump to step S134;
step S132, if the singular intermediate class is eliminatedAnd is also provided withThen->Jump to step S134;
step S133, ifOrder in principleWill->According to the->Sequencing the number of internal samples of the sub-set to obtain a sequenced set +.>And (3) removing +.>Middle frontA subset of individuals, get the set->Let->；
Step S134, save collection。
In this embodiment, the singular intermediate class is excluded according to the exclusion threshold of the singular intermediate class, so that the accuracy of extracting the central point features can be improved.
For ease of understanding by those skilled in the art, a set of preferred embodiments are provided below:
1. and (5) inputting data.
Acquiring a seed dataset for extracting a sampleThe data of the ith seed is +.>Data->Can be one-dimensional or multidimensional, and in addition, the number of the extracted point characteristic samples is also required to be provided>。
2. And (5) aggregate classification.
From the acquired seed data setAnd extract the number of point feature samples +.>The fuzzy clustering algorithm is adopted, and the aim of the step is to add according to the number of the extracted point characteristic samples>Clustering the seed data set to obtain +.>A subclass of individuals. The parameters of the fuzzy clustering algorithm FCM are set as follows:
The maximum number of iterations of FCM algorithm is nmax=100;
membership m=2;
the norm threshold of the difference of the central vector matrix of the clusters is epsilon=0.001;
the number of clusters is the number of extracted point feature samples；
Obtaining clustered datasets, wherein ,/>Indicating the i-th subset (subclass), NF is the number of subclasses after clustering, ++>。
3. And (5) sorting subclasses.
Because of the irrational nature of the number of pre-set extracted point feature samples and the large randomness of the provided seed data set, the clustered subclasses can have the following two singular subclasses: the first is a singular small subclass, with very few subclass samples, which may be single digit or even 0; the second is the singular large subclass: the number of samples of the subclass is extremely large, possibly accounting for more than 50% of the data set. For the problems presented above, further processing of the sub-classes is required. The method comprises the following steps:
3.1 parameter setting and calculation.
Presetting a lower threshold value of the lowest sample number in each subclass；
For clustered datasetsCalculating the set of sample numbers in each subclass of the clustered dataset +.>The method comprises the following steps:
wherein ,the representation is->Operators of the number of elements in the subclass;
calculation of25% quantile- >：
wherein ,representing a quantile extraction operator;
calculation of75% quantile->：
；
Calculation ofMean point of->：
；
According to and />Calculating the judging threshold value of singular large subclass +.>The method comprises the following steps:
；
according to and />Calculating the judgment threshold value of the singular small subclass +.>The method comprises the following steps: />
。
And 3.2, extracting the singular large subclass, the singular small subclass and the singular intermediate class from the set according to the judging threshold of the singular large subclass and the judging threshold of the singular small subclass.
Data sets after clusteringIn the middle, if->Then the i-th subclass->Is judged as singular large subclasses and each set of singular large subclasses is expressed as +.>, wherein ,/>Representation set->The number of subclasses of (3);
data sets after clusteringIn the middle, if->Then the i-th subclass->Is judged as singular small subclasses and each set of singular small subclasses is expressed as +.>, wherein ,/>Representation set->The number of subclasses of (3);
data sets after clusteringIn the middle, if->And->Then the i-th subclass->Is judged as singular intermediate classes and each set of singular intermediate classes is expressed as, wherein ,/>Representation set->Is a subclass of (c).
3.3 calculating intermediate operating parameters.
For collectionsThe ith subclass of (2)>Calculator class->Divided into +.>：
wherein ,representing a down-rounding operator;
The number of cuts per subclassMake up the collection->；
Counting the number of deletions as, wherein ,/>Representing the number of samples of the extracted point feature,representing the number of subclasses in the clustered dataset;
calculating the intermediate quantity value as；
Calculating a cutting threshold of the singular large subclassThe method comprises the following steps:
wherein ,represents a summation operator;
calculating an exclusion threshold for singular subclassesThe method comprises the following steps:
；
calculating an exclusion threshold for singular intermediate classesThe method comprises the following steps:
。
3.4 split singular large subclasses.
Pair aggregationCutting each subclass in (3), if the cutting threshold of the singular large subclass is +.>The segmentation is performed by the following steps: />
Step S111, pair aggregationSet i as set +.>Sequence number of each subclass in (a) and +.>；
Step S112, ifGet the collection->Is>Individual subset->And obtain subset->Data dimension +.>；
Step S113, calculating a subsetIs>Distance of 90% quantiles from 10% quantiles of the individual dimension data；
Step S114, obtaining the maximum distance value as，/>Representation->Corresponding dimension serial numbers;
step S115, at the firstIn dimension, sorting is carried out according to the data size, and the sorted set is that；
Step S116, forOne-dimensional data is added to the one-dimensional data before ordering>The data of the same column are combined in the column direction to form a new +. >Dimension data set +.>；
Step S117, according to the collectionThe corresponding division number->For->Cutting sequentially to obtain +.>, wherein ,/>Representation->And let +.>Jump to step S112;
step S118, ifThe cut data is collected as；
Cutting threshold of singular large subclassThen->, wherein ,/>Representing the absolute value taking operator.
3.5 exclude singular subclasses.
Step S121, if the exclusion threshold of the singular subclassLet->Jump to step S125;
step S122, eliminating threshold value of singular small subclassAnd is also provided withThen->Jumping to step S124;
step S123, ifOrder in principleWill->According to the->Sequencing the number of internal samples of the sub-set to obtain a sequenced set +.>And (3) removing +.>Front middle>A subset of individuals, get the set->，/>Representation->The number of subsets to be excluded; />
Step S124, collectingThe empty subset of the Chinese is excluded to obtain the set +.>；
Step S125, save set。
3.6 exclude singular intermediaries.
Step S131, if the singular intermediate class is excluded from the thresholdLet->Jump to step S134;
step S132, if the singular intermediate class is eliminatedAnd is also provided withThen->Jump to step S134;
Step S133, ifOrder in principleWill->According to the->Sequencing the number of internal samples of the sub-set to obtain a sequenced set +.>And (3) removing +.>Middle frontA subset of individuals, get the set->Let->；
Step S134, save collection。
3.7 combining the collated data.
At the level of subset, the subset is assembled、/> and />Combining to obtain a set. Within the set, ++>Is the ith subset.
4. Sub-class center extraction.
Pair aggregationThe feature extraction of the central point is carried out on each subset in the list, specifically:
for collectionsThe ith subset of +.>The median of the data of each dimension is taken to obtain a median vector +.>. Median vector is found for each subclass and combined into set +.>. AggregationThat is, a collection of extracted center point features, each member within which is a specific sample point data.
5. And outputting data.
Output data setAnd the number NN of the actually calculated central point features,/->。
Referring to fig. 2, the embodiment of the invention further provides a seed cluster-based sub-class center point feature extraction system, which includes a data acquisition unit 100, a data clustering unit 200, a sub-class dividing unit 300, a threshold calculation unit 400, a sub-class cutting unit 500, a first exclusion unit 600, a second exclusion unit 700, and a feature extraction unit 800, wherein:
A data acquisition unit 100 for acquiring a seed data set and the number of extracted point feature samples;
the data clustering unit 200 is configured to cluster the seed data set according to the number of the extracted point feature samples, and obtain a clustered data set;
the subclass dividing unit 300 is configured to divide the clustered dataset into a singular large subclass, a singular small subclass and a singular middle class;
a threshold calculating unit 400, configured to calculate a cutting threshold of the singular large subclass, an exclusion threshold of the singular small subclass, and an exclusion threshold of the singular intermediate class according to the set of the singular large subclass, the singular small subclass, and the singular intermediate class;
a subclass cutting unit 500, configured to cut the singular large subclass according to a cutting threshold of the singular large subclass, to obtain a first data set;
a first excluding unit 600, configured to exclude the small singular subclasses according to an exclusion threshold of the small singular subclasses, to obtain a second data set;
a second excluding unit 700, configured to exclude the singular intermediate class according to an exclusion threshold of the singular intermediate class, so as to obtain a third dataset;
the feature extraction unit 800 is configured to combine the first data set, the second data set, and the third data set to obtain a combined data set, and extract a center point feature corresponding to each sub-class in the combined data set.
It should be noted that, since a sub-class center point feature extraction system based on seed clustering in the present embodiment and the above-mentioned sub-class center point feature extraction method based on seed clustering are based on the same inventive concept, the corresponding content in the method embodiment is also applicable to the present system embodiment, and will not be described in detail herein.
The embodiment of the invention also provides a subclass center point feature extraction device based on seed clustering, which comprises: at least one control processor and a memory for communication connection with the at least one control processor.
The memory, as a non-transitory computer readable storage medium, may be used to store non-transitory software programs as well as non-transitory computer executable programs. In addition, the memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one magnetic disk storage device, flash memory device, or other non-transitory solid state storage device. In some embodiments, the memory optionally includes memory remotely located relative to the processor, the remote memory being connectable to the processor through a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
A non-transitory software program and instructions required to implement the seed-cluster-based sub-class center point feature extraction method of the above embodiments are stored in a memory, and when executed by a processor, one of the seed-cluster-based sub-class center point feature extraction methods of the above embodiments is performed, for example, the method steps S100 to S800 in fig. 1 described above are performed.
The system embodiments described above are merely illustrative, in that the units illustrated as separate components may or may not be physically separate, i.e., may be located in one place, or may be distributed over a plurality of network elements. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
Embodiments of the present invention also provide a computer-readable storage medium storing computer-executable instructions that are executed by one or more control processors to cause the one or more control processors to perform a seed-cluster-based sub-class center feature extraction method in the above method embodiments, for example, to perform the functions of the method steps S100 to S800 in fig. 1 described above.
Those of ordinary skill in the art will appreciate that all or some of the steps, systems, and methods disclosed above may be implemented as software, firmware, hardware, and suitable combinations thereof. Some or all of the physical components may be implemented as software executed by a processor, such as a central processing unit, digital signal processor, or microprocessor, or as hardware, or as an integrated circuit, such as an application specific integrated circuit. Such software may be distributed on computer readable media, which may include computer storage media (or non-transitory media) and communication media (or transitory media). The term computer storage media includes both volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data, as known to those skilled in the art. Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital Versatile Disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium which can be used to store the desired information and which can be accessed by a computer. Furthermore, as is well known to those of ordinary skill in the art, communication media typically embodies computer readable instructions, data structures, program modules or other data in a modulated data signal such as a carrier wave or other transport mechanism and includes any information delivery media.
While the preferred embodiments of the present application have been described in detail, the embodiments of the present application are not limited to the above-described embodiments, and those skilled in the art can make various equivalent modifications or substitutions without departing from the spirit of the embodiments of the present application, and these equivalent modifications or substitutions are included in the scope of the embodiments of the present application as defined in the appended claims.
Claims (8)
1. The subclass center point feature extraction method based on the seed clustering is characterized by comprising the following steps of:
acquiring a seed data set and the number of extracted point characteristic samples;
clustering the seed data set according to the number of the extracted point characteristic samples to obtain a clustered data set;
dividing the clustered data set into singular large subclasses, singular small subclasses and singular intermediate classes; wherein:
before the clustered data set is divided into the singular large subclass, the singular small subclass and the singular middle subclass, the subclass center point feature extraction method based on seed clustering further comprises the following steps:
presetting a threshold value of the minimum sample number in each subclass；
Computing a set of sample numbers within each sub-class of the clustered dataset The method comprises the following steps:
wherein ,the representation is->Operators of the number of elements in the subclass;
calculation of25% quantile->：
wherein ,representing a quantile extraction operator;
calculation of75% quantile->：
；
Calculation ofMean point of->：
；
Calculating the judging threshold value of the singular large subclassThe method comprises the following steps: />；
Calculating the judging threshold value of the singular small subclassThe method comprises the following steps: />；
In the clustered dataset, ifThen the i-th subclass->Is judged as singular large subclasses and each set of singular large subclasses is expressed as +.>, wherein ,/>Representation set->The number of subclasses of (3);
in the clustered dataset, ifThen the i-th subclass->Is judged as singular small subclasses and each set of singular small subclasses is expressed as +.>, wherein ,/>Representation set->The number of subclasses of (3);
in the clustered dataset, ifAnd->Then the ith subclassIs judged as singular intermediate classes and each set of singular intermediate classes is expressed as, wherein ,/>Representation set->The number of subclasses of (3);
according to the set of the singular large subclass, the singular small subclass and the singular intermediate class, calculating a cutting threshold of the singular large subclass, an exclusion threshold of the singular small subclass and an exclusion threshold of the singular intermediate class;
Cutting the singular large subclass according to a cutting threshold value of the singular large subclass to obtain a first data set;
removing the singular small subclasses according to the removal threshold value of the singular small subclasses to obtain a second data set;
removing the singular intermediate class according to the removal threshold value of the singular intermediate class to obtain a third data set;
and merging the first data set, the second data set and the third data set to obtain a merged data set, and extracting the center point characteristic corresponding to each subclass in the merged data set.
2. The seed-cluster-based sub-class center feature extraction method according to claim 1, wherein the calculating the cutting threshold of the singular large sub-class, the exclusion threshold of the singular small sub-class, and the exclusion threshold of the singular intermediate class from the set of the singular large sub-class, the singular small sub-class, and the singular intermediate class comprises:
for the setThe ith subclass of (2)>Calculating the subclass +.>Divided into +.>：
wherein ,representing a down-rounding operator;
the number of cuts per subclassMake up the collection->；
Counting the number of deletions as, wherein ,/>Representing the number of extracted point feature samples, +.>Representing the number of subclasses in the clustered dataset;
calculating the intermediate quantity value as；
Calculating a cutting threshold of the singular large subclassThe method comprises the following steps:
wherein ,represents a summation operator;
calculating an exclusion threshold for the singular subclassThe method comprises the following steps:
；
calculating the singular intermediate classExclusion thresholdThe method comprises the following steps:
。
3. the seed-cluster-based sub-class center point feature extraction method of claim 2, wherein the cutting the singular macro sub-class according to the cutting threshold of the singular macro sub-class to obtain a first data set comprises:
for the setCutting each subclass of (3) if the cutting threshold of the singular large subclass is +.>The segmentation is performed by the following steps:
step S111, for the setSetting i as the set +.>Sequence number of each subclass in (a), and；
step S112, ifTaking the set->Is>Individual subset->And obtain the subset->Data dimension +.>；
Step S113, calculating the subsetIs>Distance of 90% quantiles from 10% quantiles of the individual dimension data；
Step S114, obtaining the maximum distance value as，/>Representation->Corresponding dimension serial numbers;
step S115, at the first In dimension, sorting is carried out according to the data size, and the sorted set is that；
Step S116, forIs added to the one-dimensional data before ordering>The data of the same column are combined in the column direction to form a new +.>Dimension data set +.>；
Step S117, according to the setThe corresponding division number->For->Cutting sequentially to obtain +.>, wherein ,/>Representation->And let +.>Jump to step S112;
step S118, ifThe cut data is gathered as +.>；
If the cutting threshold of the singular large subclassThen->, wherein ,/>The representation takes absolute value.
4. The seed-cluster-based sub-class center point feature extraction method of claim 2, wherein the excluding the singular sub-class according to the exclusion threshold of the singular sub-class, obtaining a second dataset, comprises:
step S121, if the exclusion threshold of the singular small subclassLet->Jump to step S125;
step S122, eliminating threshold value of the singular small subclassAnd is also provided withThen->Jumping to step S124;
step S123, ifOrder in principleWill->According to the set->Sequencing the number of internal samples of the sub-set to obtain a sequenced set +. >And (3) removing +.>Front middle>A subset of individuals, get the set->，/>Representation->Excluding the number of subsets;
step S124, collecting the setThe empty subset of the Chinese is excluded to obtain the set +.>；
Step S125, save set。
5. The seed-cluster-based sub-class center feature extraction method of claim 2, wherein the excluding the singular intermediate class according to the exclusion threshold of the singular intermediate class, obtaining a third dataset, comprises:
step S131, if the singular intermediate class is eliminatedLet->Jump to step S134;
step S132, if the singular intermediate class is eliminatedAnd is also provided withThen->Jump to step S134;
step S133, ifOrder in principleWill->According to the set->Sequencing the number of internal samples of the sub-set to obtain a sequenced set +.>And (3) removing +.>Middle frontA subset of individuals, get the set->Let->；
Step S134, save collection。
6. A seed-cluster-based sub-class center point feature extraction system, the seed-cluster-based sub-class center point feature extraction system comprising:
the data acquisition unit is used for acquiring a seed data set and the number of the extracted point characteristic samples;
The data clustering unit is used for clustering the seed data set according to the number of the extracted point characteristic samples to obtain a clustered data set;
the subclass dividing unit is used for dividing the clustered data set into a singular large subclass, a singular small subclass and a singular middle class; wherein:
before the clustered data set is divided into the singular large subclass, the singular small subclass and the singular middle subclass, the subclass center point feature extraction method based on seed clustering further comprises the following steps:
presetting a threshold value of the minimum sample number in each subclass；
Computing a set of sample numbers within each sub-class of the clustered datasetThe method comprises the following steps:
wherein ,the representation is->Operators of the number of elements in the subclass;
calculation of25% quantile->：
wherein ,representing a quantile extraction operator;
calculation of75% quantile->：
；
Calculation ofMean point of->：
；
Calculating the judging threshold value of the singular large subclassThe method comprises the following steps: />；
Calculating the judging threshold value of the singular small subclassThe method comprises the following steps: />；
In the clustered dataset, ifThen the i-th subclass->Is judged as singular large subclasses and each set of singular large subclasses is expressed as +.>, wherein ,/>Representation set->The number of subclasses of (3);
In the clustered dataset, ifThen the i-th subclass->Is judged as singular small subclasses and each set of singular small subclasses is expressed as +.>, wherein ,/>Representation set->The number of subclasses of (3);
in the clustered dataset, ifAnd->Then the ith subclassIs judged as singular intermediate classes and each set of singular intermediate classes is expressed as, wherein ,/>Representation set->The number of subclasses of (3);
the threshold calculating unit is used for calculating a cutting threshold of the singular large subclass, an exclusion threshold of the singular small subclass and an exclusion threshold of the singular intermediate class according to the set of the singular large subclass, the singular small subclass and the singular intermediate class;
the subclass cutting unit is used for cutting the singular large subclass according to a cutting threshold value of the singular large subclass to obtain a first data set;
the first excluding unit is used for excluding the singular small subclasses according to the excluding threshold value of the singular small subclasses to obtain a second data set;
the second excluding unit is used for excluding the singular intermediate class according to the excluding threshold value of the singular intermediate class to obtain a third data set;
and the feature extraction unit is used for merging the first data set, the second data set and the third data set to obtain a merged data set, and extracting the center point feature corresponding to each subclass in the merged data set.
7. A seed cluster-based sub-class center point feature extraction device comprising at least one control processor and a memory for communicative connection with the at least one control processor; the memory stores instructions executable by the at least one control processor to enable the at least one control processor to perform the seed cluster based sub-class center feature extraction method of any one of claims 1 to 5.
8. A computer-readable storage medium storing computer-executable instructions for causing a computer to perform the seed cluster-based sub-class center feature extraction method of any one of claims 1 to 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310048970.8A CN116522183B (en) | 2023-02-01 | 2023-02-01 | Sub-class center point feature extraction method, system and equipment based on seed clustering |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310048970.8A CN116522183B (en) | 2023-02-01 | 2023-02-01 | Sub-class center point feature extraction method, system and equipment based on seed clustering |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116522183A CN116522183A (en) | 2023-08-01 |
| CN116522183B true CN116522183B (en) | 2023-09-15 |
Family
ID=87392804
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310048970.8A Active CN116522183B (en) | 2023-02-01 | 2023-02-01 | Sub-class center point feature extraction method, system and equipment based on seed clustering |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116522183B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2002095533A2 (en) * | 2001-05-18 | 2002-11-28 | Biowulf Technologies, Llc | Model selection for cluster data analysis |
| CN107368844A (en) * | 2017-06-02 | 2017-11-21 | 西安电子科技大学 | A kind of bidirectional clustering method based on AP clusters and ISA bidirectional clusterings |
| CN108734221A (en) * | 2018-05-23 | 2018-11-02 | 中国地质大学（武汉） | The improvement density peaks clustering method and system automatically determined based on subclass numbers |
| CN115204323A (en) * | 2022-09-16 | 2022-10-18 | 华智生物技术有限公司 | Seed multi-feature based clustering and synthesis method, system, device and medium |
| EP4094201A1 (en) * | 2020-02-17 | 2022-11-30 | Huawei Technologies Co., Ltd. | Method of training an image classification model |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US11709917B2 (en) * | 2020-05-05 | 2023-07-25 | Nanjing University | Point-set kernel clustering |
-
2023
- 2023-02-01 CN CN202310048970.8A patent/CN116522183B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2002095533A2 (en) * | 2001-05-18 | 2002-11-28 | Biowulf Technologies, Llc | Model selection for cluster data analysis |
| CN107368844A (en) * | 2017-06-02 | 2017-11-21 | 西安电子科技大学 | A kind of bidirectional clustering method based on AP clusters and ISA bidirectional clusterings |
| CN108734221A (en) * | 2018-05-23 | 2018-11-02 | 中国地质大学（武汉） | The improvement density peaks clustering method and system automatically determined based on subclass numbers |
| EP4094201A1 (en) * | 2020-02-17 | 2022-11-30 | Huawei Technologies Co., Ltd. | Method of training an image classification model |
| CN115204323A (en) * | 2022-09-16 | 2022-10-18 | 华智生物技术有限公司 | Seed multi-feature based clustering and synthesis method, system, device and medium |
Non-Patent Citations (1)
| Title |
|---|
| Discriminant Subclass-center Manifold Preserving Projection For Face Feature Extraction;Chao Lan et al.;《IEEE International Conference on Image Processing》;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116522183A (en) | 2023-08-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110830450A (en) | Abnormal flow monitoring method, device and equipment based on statistics and storage medium | |
| CN115392408B (en) | Method and system for detecting abnormal operation of electronic tablet counting machine | |
| CN117132778B (en) | Spectrum measurement correction calculation method and system | |
| CN117077044B (en) | Method and device for judging faults of vacuum circuit breaker for generator | |
| WO2022042152A1 (en) | Method and device for analyzing association rule of multi-dimensional network indexes, and storage medium | |
| CN106599138A (en) | Variety identification method for electrical appliances | |
| CN116308963B (en) | Government affair data analysis method and system | |
| CN117113235A (en) | Cloud computing data center energy consumption optimization method and system | |
| CN115952067A (en) | Database operation abnormal behavior detection method and readable storage medium | |
| CN116522183B (en) | Sub-class center point feature extraction method, system and equipment based on seed clustering | |
| CN109389172B (en) | Radio signal data clustering method based on non-parameter grid | |
| CN117473351B (en) | Power supply information remote transmission system based on Internet of things | |
| CN107480426A (en) | From iteration case history archive cluster analysis system | |
| CN115204323B (en) | Seed multi-feature based clustering and synthesis method, system, device and medium | |
| CN116975672B (en) | Temperature monitoring method and system for coal mine belt conveying motor | |
| CN113127464A (en) | Agricultural big data environment feature processing method and device and electronic equipment | |
| CN108415958B (en) | Weight processing method and device for index weight VLAD features | |
| CN117170979A (en) | Energy consumption data processing method, system, equipment and medium for large-scale equipment | |
| CN116719714A (en) | Training method and corresponding device for screening model of test case | |
| CN116229330A (en) | Method, system, electronic equipment and storage medium for determining video effective frames | |
| CN116912265A (en) | Remote sensing image segmentation method and system | |
| CN114610234A (en) | Storage system parameter recommendation method and related device | |
| CN113936157A (en) | Abnormal information processing method and device, storage medium and electronic device | |
| CN117992808B (en) | Big data-based computer network management method and system | |
| CN113722673A (en) | Method for optimizing meteorological data through repeated observation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |